Large-scale World Wide Web (“Web” or “WWW”) sites typically utilize large clusters, or “farms,” of Web servers to respond to incoming Web page requests. Because the volume of requests received at popular Web sites can be enormous, literally thousands of separate Web server computers may be utilized in such a Web server farm. When a request is received for a Web page available through the Web server farm, the request is assigned to one of the available Web server computers through the use of a load balancing device. By receiving Web page requests at a central location and then routing the request to an available Web server computer in this manner, a large volume of requests may be easily handled. However, while server farms configured in this manner can respond to a large number of Web page requests efficiently, the configuration and maintenance of thousands of Web server computers can be extremely time consuming and expensive.
One of the difficulties in configuring a large group of Web server computers arises from the fact that each server computer utilizes unique configuration files. For instance, the location of a particular server computer in the network may cause its configuration files to be different from the configuration files on other server computers located at other locations in the network. If a server computer is moved to a different location within the network, it may not function properly in its new location because the configuration files correspond to the previous location of the computer. Such a problem can be extremely difficult to diagnose when thousands of server computers are utilized.
Another difficulty that arises when a large-scale farm of Web servers is utilized occurs when it is necessary to “brand” the Web site for different business partners. Because the domain name to which a particular Web server responds is fixed in the configuration file, the Web server can only respond to requests from that specific domain. For instance, a Web server for providing Web-based e-mail services that is configured to receive requests for the “www.hotmail.com” domain can only respond to requests for that particular domain. If a business partner wanted to provide a branded Web-based e-mail service using the same functionality at another domain, an entirely separate Web server computer would have to be utilized to provide the branded Web site. Therefore, much of the capability of the Web server would have to be recreated for the branded Web site.
Accordingly, in light of the above problems, there is a need for a method and system for dynamically configuring a server computer that can utilize a single configuration file to dynamically configure multiple server computers. There is a further need for a method and system for dynamically configuring a server computer that allows a server computer to correctly configure itself regardless of its location within a network. There is a further need for method and system for dynamically configuring a server computer that allows a server computer to respond to requests received from multiple domains.